From OntologPSMW

Jump to: navigation, search
[ ]

Contents

Ontology Summit 2016 Communiqué - Thu 2016-04-28     (1)

Abstract     (1B)

In this session, we will discuss the dominant themes that have been emerging from the previous Track Sessions and prepare the Summit Communiqué.     (1B1)

Agenda     (1C)

  • Dial-in:     (1D4)
    • Phone (US): +1 (425) 440-5100 ... (long distance cost may apply)     (1D4A)
    • Skype: join.conference (i.e. make a skype call to the contact with skypeID="join.conference") ... (generally free-of-charge, when connecting from your computer ... ref.)     (1D4B)
      • when prompted enter Conference ID: 843758#     (1D4B1)
      • Unfamiliar with how to do this on Skype? ...     (1D4B2)
        • Add the contact "join.conference" to your skype contact list first. To participate in the teleconference, make a skype call to "join.conference", then open the dial pad (see platform-specific instructions below) and enter the Conference ID: 843758# when prompted.     (1D4B2A)
      • Can't find Skype Dial pad? ...     (1D4B3)
        • for Windows Skype users: Can't find Skype Dial pad? ... it's under the "Call" dropdown menu as "Show Dial pad"     (1D4B3A)
        • for Linux Skype users: please note that the dial-pad is only available on v4.1 (or later; or on the earlier Skype versions 2.x,) if the dialpad button is not shown in the call window you need to press the "d" hotkey to enable it. ... (ref.)     (1D4B3B)
    • instructions: once you got access to the page, click on the "settings" button, and identify yourself (by modifying the Name field from "anonymous" to your real name, like "JaneDoe").     (1D5A)
    • You can indicate that you want to ask a question verbally by clicking on the "hand" button, and wait for the moderator to call on you; or, type and send your question into the chat window at the bottom of the screen.     (1D5B)
    • thanks to the soaphub.org folks, one can now use a jabber/xmpp client (e.g. gtalk) to join this chatroom. Just add the room as a buddy - (in our case here) summit_20160428@soaphub.org ... Handy for mobile devices!     (1D5C)
  • Discussions and Q & A:     (1D6)
    • Nominally, when a presentation is in progress, the moderator will mute everyone, except for the speaker.     (1D6A)
    • To un-mute, press "*7" ... To mute, press "*6" (please mute your phone, especially if you are in a noisy surrounding, or if you are introducing noise, echoes, etc. into the conference line.)     (1D6B)
    • we will usually save all questions and discussions till after all presentations are through. You are encouraged to jot down questions onto the chat-area in the mean time (that way, they get documented; and you might even get some answers in the interim, through the chat.)     (1D6C)
    • During the Q&A / discussion segment (when everyone is muted), If you want to speak or have questions or remarks to make, please raise your hand (virtually) by clicking on the "hand button" (lower right) on the chat session page. You may speak when acknowledged by the session moderator (again, press "*7" on your phone to un-mute). Test your voice and introduce yourself first before proceeding with your remarks, please. (Please remember to click on the "hand button" again (to lower your hand) and press "*6" on your phone to mute yourself after you are done speaking.)     (1D6D)
  • RSVP to gruninger@mie.utoronto.ca with your affiliation appreciated, ... or simply just by adding yourself to the "Expected Attendee" list below (if you are a member of the community already.)     (1D8)
  • This session, like all other Ontolog events, is open to the public. Information relating to this session is shared on this wiki page.     (1D9)
  • Please note that this session may be recorded, and if so, the audio archive is expected to be made available as open content, along with the proceedings of the call to our community membership and the public at-large under our prevailing open IPR policy.     (1D10)

Attendees     (1E)

Proceedings     (1F)

[12:27] Gary Berg-Cross: Hi Brandon. Glad you could make this. I think that you'll have much to contribute.     (1F1)

[12:28] brandon whitehead: Hi Gary. thanks. I've been out and on holiday...and now have to catch up with all the email conversations. Looks great so far!     (1F2)

[12:29] Russ: Could someone post the link for the slides please     (1F3)

[12:36] ToddSchneider:     (1F5)

Semantic Interoperability Taxonomic Subdimension Structure     (1F6)

Operational Context Subdimensions     (1F10)

Global Versus Specialization of Domain Knowledge for Communities of Interest     (1F20)

[12:36] ToddSchneider: The previous post is a taxonomy from the SCOPE model document, version 1.     (1F21)

[12:37] Mark Underwood: Shortcut tothe registration page: http://bit.ly/1YVJ61X     (1F22)

[12:40] ToddSchneider: What about a problem statement?     (1F23)

[12:43] MichaelGruninger: A complementary idea for the overall Communique organization is     (1F24)

[12:43] MichaelGruninger:     (1F25)

1. The Story Thus Far ... This section will pose the problem of semantic integration and summarize how previous Summits have touched upon some of the key ideas     (1F26)

2. Where Are We Now? This section will include most of the material that summarizes the contributions specific to this Summit     (1F27)

3. Where Are We Going? This includes recommendations and possibly forecasts     (1F28)

[12:44] ToddSchneider: It was asked whether semantic integration of systems reducible to ontology integration? No. Even if two systems conformed to the same interpretation of terms used throughout, there can be many other implementation and technology specific (and perhaps non-technical) differences to impede to systems integration. Some of these may network security, access control, data formats, version differences (among operating systems or software), intellectual property concerns, communication constraints, or budgeting priorities.     (1F29)

[12:45] MichaelGruninger: All of today's material is on the Session page: http://ontologforum.org/index.php/ConferenceCall_2016_04_28     (1F30)

[12:46] ToddSchneider: So, is one requirement of semantic integration to be able to integrate, or integrate with, existing standards?     (1F31)

[12:47] Mark Underwood: If you email to me I can post it to the other site     (1F32)

[12:47] ToddSchneider: Michael, is it out of scope to have a summary of previous attempts at 'semantic integration'?     (1F33)

[12:47] Gary Berg-Cross: Our current summary is at http://ontologforum.org/index.php/Talk:ConferenceCall_2016_03_31     (1F34)

[12:47] brandon whitehead: @Todd, I think integrating with existing standards is not a *requirement* per se, but it is certainly low hanging fruit.     (1F35)

[12:47] Mark Underwood: Todd's draft for the Engineering/Cloud Track is here: http://ontologysummit.org/wp-content/uploads/2016/04/EngineeringSynthesis-Draft1_28Apr2016-mau.pdf     (1F36)

[12:49] ToddSchneider: Brandon, in the area of manufacturing there are many very detailed specifications for common components, e.g., fasteners. These are the sorts of 'standards' I had in mind. Perhaps I should have used the term 'specification' instead.     (1F37)

[12:50] AndreaWesterinen: @Todd, In terms of using standards as "reference ontologies", like in finance, then it is also necessary to integrate them into the domain ontologies. That was the purpose of Rebecca's and my talk last week.     (1F38)

[12:50] ToddSchneider: Brandon, if a semantic integration solution doesn't integrate with 'standards' or 'specifications' that are currently in use, I expect little acceptance.     (1F39)

[12:50] RaviSharma: Mark - url for the Other site?     (1F40)

[12:50] LeoObrst: Whats needed:     (1F41)

Multiple contexts, views, application & user perspectives     (1F42)

Multiple levels of precision, specification, definiteness required     (1F43)

Multiple levels of semantic model verisimilitude, fidelity, granularity, dynamicity     (1F44)

Multiple kinds of semantic mappings, transformations needed:     (1F45)

  • Entities, Relations, Properties, Ontologies, Model Modules, Namespaces, Meta-Levels, Facets (i.e., properties of properties), Units of Measure, Conversions, Theories, Interpretations, etc.     (1F46)
  • Upper/Foundational, Mid-level, and Utility Ontologies are important to be able to interrelate domain ontologies.     (1F47)

[12:51] ToddSchneider: Andrea, so you agree there is a requirement for semantic integration?     (1F48)

[12:51] Mark Underwood: Hi Ravi: ontologysummit.org     (1F49)

[12:51] brandon whitehead: Todd, I see. thank you. That does make sense. I'm speaking from (primarily) a geosciences perspective. Geoscience communities are, in my view, lacking standards (and certainly specifications).     (1F50)

[12:52] ToddSchneider: Leo, look at the list of notions (i.e., the taxonomy) I posted earlier. Semantic interoperability has many 'dimensions'.     (1F51)

[12:52] Mark Underwood: Cross-cutting remarks: Some backlinks to previous summit content     (1F52)

[12:52] AndreaWesterinen: @Todd, Yes, I do.     (1F53)

[12:53] ToddSchneider: Brandon, Gary addresses the lack or looseness of specification in the geo-science area.     (1F54)

[12:53] LeoObrst: Also, I suggest we look at an old paper by Steve Ray, which he's approved our use of, if desired. Perhaps we can post that too?     (1F55)

[12:54] brandon whitehead: Todd, yes. I think Gary and I are on the same page.     (1F57)

[12:54] brandon whitehead: (at least the same book)  :)     (1F58)

[12:54] Gary Berg-Cross: One point on Upper Level ontologies which were mentioned as part of the Health Science area is that the Upper level Os are too complex for many and hard to understand. Hence a move in the modular direction.     (1F59)

[12:57] Gary Berg-Cross: Todd was it "Semantic Interoperability Conceptual Framework (SICF) "?     (1F60)

[12:57] Mark Underwood: A speaker referred to this W3C undertaking - not sure what current status is: http://www.automotive-ontology.org/     (1F61)

[12:58] Mark Underwood: It was referenced on https://www.w3.org/wiki/WebSchemas/Vehicles     (1F62)

[12:58] AndreaWesterinen: I would say that many people used (past tense) WSDL, I see much more use of RESTful interfaces now.     (1F63)

[12:58] AndreaWesterinen: For simplicity reasons, etc.     (1F64)

[12:58] Mark Underwood: Ravi, true, but is it with ontologies?     (1F65)

[12:59] Donna Fritzsche: great input Ravi regarding WSDL/SOa- lets have additional discussion on the listserve or chat. Thank-you!     (1F66)

[13:00] Donna Fritzsche: Action item - to see how that intersects with SI.     (1F67)

[13:00] Mark Underwood: Big uptake for WSDL, but little if anything that would be of interest for our CoI     (1F68)

[13:00] Mark Underwood: -iMHO     (1F69)

[13:02] ChristiKapp: Don't ReSTful interfaces use XML & Json? Wouldn't those interfaces would benefit by being able to hook into industry specific ontologies somehow instead of re-inventing the structure with each service definition?     (1F70)

[13:02] Mark Underwood: @Todd, @Andrea - I'll be adding content to Todd's draft to cover RESTful, APIs etc, which was what I had hoped we sould expose. I'm hoping to have a F2F presentation by iServe which registers their code in Github & follows the cloud services de facto design pattern     (1F71)

[13:03] Gary Berg-Cross: I think that our Communioque should have a clear and early section that discusses what semantic interoperability and integration mean and how they relate.     (1F72)

[13:04] Mark Underwood: @Christi. In short, yes.     (1F73)

[13:07] Gary Berg-Cross: Along with the reference ontology discussion we should mention Deeper Semantics which Brandon brought up along with Torsten & K Janowicz.     (1F74)

[13:08] Gary Berg-Cross: Torsten isn't on the call (yet) but he recently posted this as part of the forum: Some thoughts about the original question of why we still have a "semantic mess". I agree that all four factors contribute, but I have another theory that strongly supports 4 (we don't have the right ontologies).     (1F75)

But from my experience with geoscience ontologies, I think a big part of the reason is that many of the geoscience terms are somewhat semi-scientific (e.g. river, channel, water body) and scientists simply don't even have a clear understanding of them. So if they don't even have a precise understanding of high-level terms (which are the ones most likely to be reused across ontologies or datasets), we have no choice but to help them develop it while we develop suitable ontologies. I think part of our job as ontologists is to give them a language (better, more "neutral" terms) to develop and refine their own understanding. I think this is what reference ontologies could provide.     (1F76)

I suspect this is happening in many other domains as well. A rare exception might be the biosciences, where the terms are much more restricted (no bioscientist would confuse a chromosome for something else). Maybe this partially explains why ontologies have been most successful in the biosciences.     (1F77)

Wondering whether others have a similar experience or would disagree.     (1F78)

[13:09] brandon whitehead: +1 Gary.     (1F79)

[13:09] RaviSharma: Andrea - please clarify accessing data through federation Vs in enterprise repositories relative costs, merits?     (1F80)

[13:10] Mark Underwood: +1 Tools for KE sparse, especially open source. There is no KE / ontology equivalent for Apache stack, e.g., Hadoop     (1F81)

[13:12] Gary Berg-Cross: Tools are needed to help find relevant concepts. as was mentioned by Andrea and part of the Sem+ talk     (1F82)

[13:12] RaviSharma: Andrea - minimum pre-requisite to SI is agreement on the core vocabulary for information sharing between domains.     (1F83)

[13:13] Gary Berg-Cross: One of the "integration" problem is that there are different languages for formalizastion which are not easily handled.     (1F84)

[13:14] Gary Berg-Cross: We clearly have enough material for a Communique and a challenge will be structuring this and communicating it well.     (1F85)

[13:16] Gary Berg-Cross: I think that the Communique should say something about the limitations of taxonomies along, which precedes the issues with top level ontologies.     (1F86)

[13:16] RaviSharma: Can we sharpen difference between concept and context - in where and how to use - to level set interoperability     (1F87)

[13:19] Gary Berg-Cross: Explaining the connection of metadata and ontologies to the Communique reader would be a useful service.     (1F88)

[13:19] AndreaWesterinen: @Ravi, I am not sure of your question. Federating data across enterprise repositories is usually necessary to just run a business. Procurement systems integrating with finance for payments, integrating with people who need/use what is procured, ...     (1F89)

[13:19] Mark Underwood: @Andrea - nice deck     (1F90)

[13:19] RaviSharma: Can we not use machine learning to level-set concepts and contexts - especially in the sense of sharing data (information) to extract useful entities candidates for SI? This solves concept, context level setting as well as defines terms or data elements interoperable (essentially core SI).     (1F91)

[13:20] AndreaWesterinen: @Ravi, Is the question about federation vs collecting all the data in one super repository?     (1F92)

[13:20] AndreaWesterinen: @Mark, Thanks.     (1F93)

[13:20] Gary Berg-Cross: RAvi this ML idea might be a research question rather than a working best practice.     (1F94)

[13:21] AndreaWesterinen: @Gary, Agree on metadata versus/with ontologies.     (1F95)

[13:21] RaviSharma: Andrea - I was referring to cross enterprise as federation not just systems integration in the enterprise.     (1F96)

[13:21] Mark Underwood: +1 for connections to metadata mgmt; MDM is commonly accepted technology for data warehouses. Ontologies aren't.     (1F97)

[13:22] Mark Underwood: Haven't we [Summit} group tried that previously     (1F98)

[13:22] RaviSharma: Add relations (not in hierarchy) and you get ontologies?     (1F99)

[13:23] AndreaWesterinen: @Ravi, You could say that compliance reporting is a form of "cross enterprise federation" (from a company to the government/audit agency).     (1F100)

[13:23] Gary Berg-Cross: Increasingly people are expressing metadata in RDF form which can then be "explained" by an ontology. So that is one connection.     (1F101)

[13:24] Mark Underwood: @Donna, I will flesh out my part of slide 9     (1F102)

[13:24] AndreaWesterinen: @Gary, +1. My customer is doing exactly that (RDF triples, explained by an ontology)     (1F103)

[13:25] RaviSharma: YES then is it costlier to send generated reports or allow regulators to reach data and themselves generate reports - compare and costs?     (1F104)

[13:25] RaviSharma: last message was for Andrea.     (1F105)

[13:26] ToddSchneider: Donna, The SCOPE model was developed under the auspices of NCOIC.     (1F106)

[13:26] AndreaWesterinen: @Ravi, I am pretty sure that most companies don't want outsiders to access their data. This could create a huge risk and allow other (unintended) analyses.     (1F107)

[13:27] ToddSchneider: Do we need to address the drivers for semantic interoperability?     (1F108)

[13:29] AndreaWesterinen: @Todd, Finance definitely has drivers - compliance, etc. But businesses similarly have this need, although they don't think of it as semantic interoperability. This is my example of a business' workers/employees, vs procurement, vs finance, ...     (1F109)

[13:30] ToddSchneider: Semantic integration or semantic interoperability?     (1F110)

[13:30] Mark Underwood: Michael reminds audience to post definitions / operationalization of semantic interop to the listserv     (1F111)

[13:30] LeoObrst: Thanks, folks, got to go.     (1F112)

[13:30] ToddSchneider: Andrea, agreed. So is the underlying driver mostly cost? Does risk come into it?     (1F113)

[13:30] Gary Berg-Cross: I believe that we can leverage some of the material on SI that Leo briefed early on. I would be willing to help on a definitional section which might add some success stories that illustrate it.     (1F114)

[13:31] AndreaWesterinen: FYI ... here is what Wikipedia says about semantic interoperability ... Semantic interoperability is the ability of computer systems to exchange data with unambiguous, shared meaning. Semantic interoperability is a requirement to enable machine computable logic, inferencing, knowledge discovery, and data federation between information systems.     (1F115)

[13:32] AndreaWesterinen: And, here is what Wikipedia says about semantic interoperation ... Semantic integration is the process of interrelating information from diverse sources, for example calendars and to do lists, email archives, presence information (physical, psychological, and social), documents of all sorts, contacts (including social graphs), search results, and advertising and marketing relevance ...     (1F116)

[13:34] Mark Underwood: Mark (recapping my comment)- I mention software engineering theme and the prevalence (+ odd status) of schema.org     (1F117)

[13:34] ToddSchneider: Semantic Interoperability issues often masquerade as technical problems.     (1F118)

[13:35] AndreaWesterinen: @Todd, I think that the finance drivers are getting good business data and inferring from it (esp inferring risks) and then compliance.     (1F119)

[13:35] ToddSchneider: Here are some questions from SCOPE training that may be helpful in creating the communique. When does semantic interoperability come into play? Where does semantic interoperability come into play? How does semantic interoperability come into play?     (1F120)

[13:36] RaviSharma: Andrea - federation among entities is common, especially for regulatory audits and risk assessments, but in industry exchanges the suppliers see the customer inventory etc online. Similarly elsewhere such as in DoD for joint developments and as part of procurement and development being one and suppliers sharing info.     (1F121)

[13:36] Russ: Where are Steve's papers available?     (1F122)

[13:37] Mark Underwood: Steve, echoing Christi, "JSON just a syntactic change"     (1F124)

[13:37] ToddSchneider: Here's a definition of semantic interoperability that we use in SCOPE: The ability for a receiver of a message to interpret the contents in a way not too dissimilar from the intended interpretation of the sender.     (1F125)

[13:37] Mark Underwood: @Steve "We haven't hit mainstream yet IMHO" - +1     (1F126)

[13:37] Gary Berg-Cross: I think that Steve's points might be added to the Communique as part of the recant environment.     (1F127)

[13:38] RaviSharma: Gary - JSON will be less formal and therefore less accuracy as compared to XML.     (1F128)

[13:38] Donna Fritzsche: agile process does not favor foundational work.     (1F129)

[13:38] ToddSchneider: Steve, but the 'O' word is occuring more frequently in job descriptions (at least in the DC Metro area).     (1F130)

[13:38] Gary Berg-Cross: Light weight approaches are what people are trying to get buy in.     (1F131)

[13:38] Donna Fritzsche: Agile process comes not just from developers but management.     (1F132)

[13:39] Donna Fritzsche: Eliza's slide had some good tips for handling this set of challenges.     (1F133)

[13:40] Mark Underwood: @Steve, no joke - that would be the value of cloud service libs that would solve "real world" problems. Developers would embrace when they solve problems IFF they work within their dev environments (e.g., Visual Studio, Eclipse), not a separate tool that requires a whole new learning experience     (1F134)

[13:40] Gary Berg-Cross: I think that the idea of standard vocabularies for metadata was what people have been trying but find less use than they hope for without formalization.     (1F135)

[13:40] RaviSharma: Gary - that was meant to be a question.     (1F136)

[13:41] Gary Berg-Cross: Yes, I understood but it might be a side issue for this conversation.     (1F137)

[13:41] Russ: Thanks Mark     (1F138)

[13:41] Donna Fritzsche: Need examples, requirements, and use cases     (1F139)

[13:42] RaviSharma: Gary - similarly RDF vs UML but RDF is being used more often as ease of usability so will JSON.     (1F140)

[13:43] Mark Underwood: FYI A 2004 The Semantic Interoperability Community of Practice (SICoP) of the Federal CIO Council slide deck http://www.daml.org/meetings/2004/12/pi/SICoP.pdf     (1F141)

[13:44] ToddSchneider: Many times problems are not recognized as being caused by inconsistent or incomplete semantics (or sufficiently constrained interpretations).     (1F142)

[13:44] Ram D. Sriram1: In terms of case studies, we had a number in the health care session. Again, in terms of ontology use I refer youto the slide that Leo Showed on the ontology spectrum.     (1F143)

[13:46] AndreaWesterinen: @Donna, For slide 18 (2014 Summit), Mike, Gary and I ran a track on reuse. That definitely comes into play.     (1F144)

[13:47] AndreaWesterinen: I could help write something for that.     (1F145)

[13:48] Mark Underwood: @Andrea +1 reuse is key     (1F146)

[13:48] Gary Berg-Cross: Sorry, I have to sign off now, but will check the site for the full chat later.     (1F147)

[13:49] Mark Underwood: ciao , Gary     (1F148)

[13:51] ToddSchneider: To focus the work, a finalized outline will help.     (1F149)

[13:51] Ram D. Sriram1: In terms of what you need from the track chairs, please send an email requesting the information you need.     (1F150)

[13:51] AndreaWesterinen: I would not want track by track details. We have some general and specific topics.     (1F151)

[13:53] AndreaWesterinen: We need to organize the concepts.     (1F152)

[13:56] RaviSharma: Donna - How relevant is domain detail in communiques, I mean details, assuming domain status summary is there.     (1F153)

[13:56] Ram D. Sriram1: @Andrea: Yes, the General Chairs organize the concepts from all the input from individual track chairs.     (1F154)

[13:58] RaviSharma: Donna - You have started well now can we provide some observations, unfortunately domain based in some cases and not necessarily ontology formalism or tools based but items that will help interoperability.     (1F155)

[13:59] SteveRay: The other paper I made reference to is at http://www.nist.gov/manuscript-publication-search.cfm?pub_id=821713     (1F156)

[13:59] Mark Underwood: @Thx steve     (1F157)

[13:59] AndreaWesterinen: Sorry, have to leave.     (1F158)

[14:00] AndreaWesterinen: Will check back later for any specific work items.     (1F159)

[14:00] Russ: I might be able to contribute. Can I get the link to the slide deck.     (1F160)

[14:01] Mark Underwood: @Steve, Leo forwarded that but it has an ASME copyright, so I didn't jump on that to distribute at ontologysummit.org     (1F161)

[14:03] Russ: Thanks Mark     (1F163)

[14:07] Mark Underwood: I can vouch for Michael's heavy lift     (1F164)

[14:10] ToddSchneider: Power Point is only good for superficial communication (and those with short attention spans).     (1F165)

[14:11] RaviSharma: Donna - yes thanks for open mind - and Ram's comments also are relevant - we need to converge progressively and make improvements and final consensus only at Summit sessions. Slide 21 is more up front to frame thoughts about use cases, status of domain uses and tools.     (1F166)

[14:12] RaviSharma: Donna and Michael - I would provide inputs for you to edit - but where would be a vesioned document?     (1F167)

[14:13] Mark Underwood: Mark @knowelngr.com     (1F168)

[14:13] Mark Underwood: typo @knowlengr.com     (1F169)

[14:13] Mark Underwood: oy knowlengr@gmail.com (3rd time a charm)     (1F170)

[14:15] SteveRay: @Mark: A reasonable conclusion regarding the other paper, but in fact the federal government asserts its right that all work published by government employees about their work is free of copyright. Some publishers slap the copyright on the papers anyway, but its not enforceable.     (1F171)

[14:16] Mark Underwood: @Steve That is good.     (1F172)

[14:18] Russ: I might be able to help with Characterization of Use Cases and Requirements     (1F173)

[14:20] Mark Underwood: Slide 21 follows:     (1F174)

[14:20] Mark Underwood:     (1F175)

Thoughts & Dialogue     (1F176)

Challenges     (1F181)

Processes & Methodologies     (1F185)

Conversational Framework     (1F190)

  • Characterization of Use Cases and Requirements     (1F191)
  • Characterization of Semantic Models (what functionality do they offer consuming applications/ecosystem participants)     (1F192)
  • Measures of Appropriate Semantic Interoperability     (1F193)
  • Conditions and Requirements for Successful SI Projects     (1F194)